We use the following packages

library(MASS)     # Datasets
library(mice)     # Boys dataset
library(dplyr)    # Data manipulation
library(magrittr) # Pipes
library(ggplot2)  # Plotting suite

New functions:

  • hist(): histogram
  • plot(): R’s plotting device
  • barplot(): bar plot function
  • boxplot(): box plot function
  • density(): function that calculates the density
  • ggplot(): ggplot’s plotting device

Why visualise?

  • We can process a lot of information quickly with our eyes
  • Plots give us information about
    • Distribution / shape
    • Irregularities
    • Assumptions
    • Intuitions
  • Summary statistics, correlations, parameters, model tests, p-values do not tell the whole story

ALWAYS plot your data!

Why visualise?

Source: Anscombe, F. J. (1973). “Graphs in Statistical Analysis”. American Statistician. 27 (1): 17–21.

What we will do

  • A few plots in base graphics in R
  • Plotting with ggplot2 graphics

Plots

Histogram

hist(boys$hgt, main = "Histogram", xlab = "Height")

Density

dens <- density(boys$hgt, na.rm = TRUE)
plot(dens, main = "Density plot", xlab = "Height", bty = "L")

Scatter plot

plot(x = boys$hgt, y = boys$wgt, main = "Scatter plot", 
     xlab = "Height", ylab = "Weight", bty = "L")

Box plot

boxplot(boys$hgt ~ boys$reg, main = "Boxplot", 
        xlab = "Region", ylab = "Height")

Box plot II

boxplot(hgt ~ reg, boys,  main = "Boxplot", xlab = "Region", 
        ylab = "Height", lwd = 2, notch = TRUE, col = rainbow(5))

A lot can be done in base R!

boys %>% md.pattern() # from mice

##     age reg wgt hgt bmi hc gen phb  tv     
## 223   1   1   1   1   1  1   1   1   1    0
## 19    1   1   1   1   1  1   1   1   0    1
## 1     1   1   1   1   1  1   1   0   1    1
## 1     1   1   1   1   1  1   0   1   0    2
## 437   1   1   1   1   1  1   0   0   0    3
## 43    1   1   1   1   1  0   0   0   0    4
## 16    1   1   1   0   0  1   0   0   0    5
## 1     1   1   1   0   0  0   0   0   0    6
## 1     1   1   0   1   0  1   0   0   0    5
## 1     1   1   0   0   0  1   1   1   1    3
## 1     1   1   0   0   0  0   1   1   1    4
## 1     1   1   0   0   0  0   0   0   0    7
## 3     1   0   1   1   1  1   0   0   0    4
##       0   3   4  20  21 46 503 503 522 1622

Many R objects have a plot() method

result <- lm(age~wgt, boys)
plot(result, which = 1)

Neat! But what if we want more control?

ggplot2

What is ggplot2?

Layered plotting based on the book The Grammer of Graphics by Leland Wilkinsons.

With ggplot2 you

  1. provide the data
  2. define how to map variables to aesthetics
  3. state which geometric object to display
  4. (optional) edit the overall theme of the plot

ggplot2 then takes care of the details

An example: scatterplot

1: Provide the data

boys %>%
  ggplot()

2: map variable to aesthetics

boys %>%
  ggplot(aes(x = age, y = bmi))

3: state which geometric object to display

boys %>%
  ggplot(aes(x = age, y = bmi)) +
  geom_point()

An example: scatterplot

Why this syntax?

Create the plot

gg <- 
  boys %>%
  ggplot(aes(x = age, y = bmi)) +
  geom_point(col = "dark green")

Add another layer (smooth fit line)

gg <- gg + 
  geom_smooth(col = "dark blue")

Give it some labels and a nice look

gg <- gg + 
  labs(x = "Age", y = "BMI", title = "BMI trend for boys") +
  theme_minimal()

Why this syntax?

plot(gg)

Why this syntax?

Aesthetics

  • x
  • y
  • size
  • colour
  • fill
  • opacity (alpha)
  • linetype
  • …

Aesthetics

gg <- 
  boys %>% 
  filter(!is.na(reg)) %>% 
  
  ggplot(aes(x      = age, 
             y      = bmi, 
             size   = hc, 
             colour = reg)) +
  
  geom_point(alpha = 0.5) +
  
  labs(title  = "BMI trend for boys",
       x      = "Age", 
       y      = "BMI", 
       size   = "Head circumference",
       colour = "Region") +
  theme_minimal()

Aesthetics

plot(gg)

Geoms

  • geom_point

  • geom_bar

  • geom_line

  • geom_smooth

  • geom_histogram

  • geom_boxplot

  • geom_density

Geoms: Bar

Geoms: Line

Geoms: Smooth

Geoms: Boxplot

Geoms: Density

Changing the Style: Themes

  • Themes determine the overall appearance of your plot
  • standard themes: e.g., theme_minimal(), theme_classic(), theme_bw(), …
  • extra libraries with additional themes: e.g., ggthemes
  • customize own theme using options of theme()

Changing the Style: Themes

Practical